R Notebook

INTRO

Importamos las librerías a utilizar

require(dplyr)

Loading required package: dplyr

Attaching package: ‘dplyr’

The following objects are masked from ‘package:stats’:

    filter, lag

The following objects are masked from ‘package:base’:

    intersect, setdiff, setequal, union

require(ggplot2)

Loading required package: ggplot2

require(tsibble)

Loading required package: tsibble
Warning: package ‘tsibble’ was built under R version 4.2.2
Attaching package: ‘tsibble’

The following objects are masked from ‘package:base’:

    intersect, setdiff, union

require(lubridate)

Loading required package: lubridate

Attaching package: ‘lubridate’

The following object is masked from ‘package:tsibble’:

    interval

The following objects are masked from ‘package:base’:

    date, intersect, setdiff, union

library(prophet)

Warning: package ‘prophet’ was built under R version 4.2.2Loading required package: Rcpp
Loading required package: rlang

EDA

Importamos los datasets con los que se va a trabajar:

df_abc <- read.csv("./datasets/202208_PAX15min-ABC.csv", sep=";")
df_deh <- read.csv("./datasets/202208_PAX15min-DEH.csv", sep=";")

df <- bind_rows(df_abc, df_deh)
rm(df_abc, df_deh)
print(colnames(df))

 [1] "FECHA"           "DESDE"           "HASTA"           "LINEA"           "MOLINETE"        "ESTACION"        "pax_pagos"       "pax_pases_pagos" "pax_franq"      
[10] "pax_TOTAL"      

Reviso datasets anteriores:

df_2022 <- read.csv("./datasets/molinetes_2022.csv", sep=",")
print(colnames(df_2022))

 [1] "FECHA"           "DESDE"           "HASTA"           "LINEA"          
 [5] "MOLINETE"        "ESTACION"        "pax_pagos"       "pax_pases_pagos"
 [9] "pax_franq"       "pax_TOTAL"      

df_2022 <- df_2022 %>%
  mutate(ts = dmy_hms(paste0(FECHA," " ,DESDE)))

df_2021 <- read.csv("./datasets/molinetes_2021.csv", sep=";")
print(colnames(df_2021))

 [1] "periodo"         "FECHA"           "DESDE"           "HASTA"          
 [5] "LINEA"           "MOLINETE"        "ESTACION"        "pax_pagos"      
 [9] "pax_pases_pagos" "pax_franq"       "pax_TOTAL"      

df_2021 <- df_2021 %>%
  mutate(ts = dmy_hms(paste0(FECHA," " ,DESDE)))

Warning:  1043613 failed to parse.

summary(df_2021)

    periodo        FECHA              DESDE              HASTA          
 Min.   :2021   Length:8071680     Length:8071680     Length:8071680    
 1st Qu.:2021   Class :character   Class :character   Class :character  
 Median :2021   Mode  :character   Mode  :character   Mode  :character  
 Mean   :2021                                                           
 3rd Qu.:2021                                                           
 Max.   :2021                                                           
                                                                        
    LINEA             MOLINETE           ESTACION           pax_pagos     
 Length:8071680     Length:8071680     Length:8071680     Min.   :  0.00  
 Class :character   Class :character   Class :character   1st Qu.:  2.00  
 Mode  :character   Mode  :character   Mode  :character   Median :  7.00  
                                                          Mean   : 11.28  
                                                          3rd Qu.: 15.00  
                                                          Max.   :292.00  
                                                                          
 pax_pases_pagos      pax_franq         pax_TOTAL    
 Min.   :0.000000   Min.   : 0.0000   Min.   :  1.0  
 1st Qu.:0.000000   1st Qu.: 0.0000   1st Qu.:  3.0  
 Median :0.000000   Median : 0.0000   Median :  7.0  
 Mean   :0.006266   Mean   : 0.6143   Mean   : 11.9  
 3rd Qu.:0.000000   3rd Qu.: 1.0000   3rd Qu.: 16.0  
 Max.   :9.000000   Max.   :54.0000   Max.   :298.0  
                                                     
       ts                        
 Min.   :2021-01-01 08:00:00.00  
 1st Qu.:2021-06-16 13:00:00.00  
 Median :2021-09-07 12:30:00.00  
 Mean   :2021-08-20 11:27:10.45  
 3rd Qu.:2021-11-06 08:15:00.00  
 Max.   :2021-12-31 22:00:00.00  
 NA's   :1043613                 

summary(df_2022)

    FECHA              DESDE              HASTA              LINEA          
 Length:2468478     Length:2468478     Length:2468478     Length:2468478    
 Class :character   Class :character   Class :character   Class :character  
 Mode  :character   Mode  :character   Mode  :character   Mode  :character  
                                                                            
                                                                            
                                                                            
                                                                            
   MOLINETE           ESTACION           pax_pagos      pax_pases_pagos   
 Length:2468478     Length:2468478     Min.   :  0.00   Min.   : 0.00000  
 Class :character   Class :character   1st Qu.:  4.00   1st Qu.: 0.00000  
 Mode  :character   Mode  :character   Median : 12.00   Median : 0.00000  
                                       Mean   : 18.72   Mean   : 0.02097  
                                       3rd Qu.: 26.00   3rd Qu.: 0.00000  
                                       Max.   :390.00   Max.   :11.00000  
                                                                          
   pax_franq        pax_TOTAL            ts                        
 Min.   : 0.000   Min.   :  1      Min.   :2022-04-01 05:15:00.00  
 1st Qu.: 0.000   1st Qu.:  4      1st Qu.:2022-04-13 16:45:00.00  
 Median : 0.000   Median : 12      Median :2022-04-26 19:15:00.00  
 Mean   : 0.917   Mean   : 20      Mean   :2022-04-28 17:18:53.03  
 3rd Qu.: 1.000   3rd Qu.: 27      3rd Qu.:2022-05-12 17:45:00.00  
 Max.   :65.000   Max.   :391      Max.   :2022-05-31 23:30:00.00  
                  NA's   :851968                                   

Grafico por línea y separo por estacion

df_linea <- df %>% 
  group_by(FECHA,DESDE,HASTA,LINEA,ESTACION) %>% 
  summarize(pasajeros = sum(pax_TOTAL)) %>% 
  mutate(ts = dmy_hms(paste0(FECHA," " ,DESDE)))%>%
  ungroup() %>% 
  select(-c(FECHA, DESDE, HASTA))

`summarise()` has grouped output by 'FECHA', 'DESDE', 'HASTA', 'LINEA'. You can override using the `.groups` argument.

library(plotly)
library(htmlwidgets)

setwd("D:\\OneDrive\\Personal\\Maestria Data Mining\\Cuatrimestre_2\\Enfoque Estadístico del Aprendizaje\\TP\\EEA-TPs\\TP2\\tp2-EEA-2022\\output")

Warning: The working directory was changed to D:/OneDrive/Personal/Maestria Data Mining/Cuatrimestre_2/Enfoque Estadístico del Aprendizaje/TP/EEA-TPs/TP2/tp2-EEA-2022/output inside a notebook chunk. The working directory will be reset when the chunk is finished running. Use the knitr root.dir option in the setup chunk to change the working directory for notebook chunks.

dia <- c("2022-08-08","2022-08-09","2022-08-10","2022-08-11","2022-08-12","2022-08-13","2022-08-14")
dia_final <- c("2022-08-09","2022-08-10","2022-08-11","2022-08-12","2022-08-13","2022-08-14","2022-08-15")
nombre_dia <- c('Lunes','Martes','Miercoles','Jueves','Viernes','Sabado','Domingo')
i<- 1

lista_lineas <- c('LineaA','LineaB','LineaC','LineaD','LineaE','LineaH')
for (dia_inicial in dia){
  print(i)
  dia <- dia_inicial
  dia1 <- dia_final[i]
  nombre <- nombre_dia[i]
  print(dia)
  print(dia1)
  print(nombre)
    for (linea in lista_lineas){
    p <- df_linea %>%
    filter(ts > dia,ts<dia1,LINEA == linea) %>%
    ggplot() + geom_line(aes(ts, pasajeros,color=ESTACION)) + ggtitle(paste('Pasajeros para el dia',nombre,'en',linea,sep=" ")+
                                                                        scale_x_date(date_breaks = "1 hour", date_labels = "%H:%M") +
                                                                        scale_x_date(date_minor_breaks = "15 minutes"))
    print(ggplotly(p))
    
    saveWidget(ggplotly(p), file = paste('Pasajeros para el dia',nombre,'en',linea,'.html',sep=" "));
    }
  i <- i+1
}

[1] 1
[1] "2022-08-08"
[1] "2022-08-09"
[1] "Lunes"

[1] 2
[1] "2022-08-09"
[1] "2022-08-10"
[1] "Martes"

[1] 3
[1] "2022-08-10"
[1] "2022-08-11"
[1] "Miercoles"

[1] 4
[1] "2022-08-11"
[1] "2022-08-12"
[1] "Jueves"

[1] 5
[1] "2022-08-12"
[1] "2022-08-13"
[1] "Viernes"

[1] 6
[1] "2022-08-13"
[1] "2022-08-14"
[1] "Sabado"

[1] 7
[1] "2022-08-14"
[1] "2022-08-15"
[1] "Domingo"

Agrupamos la información a nivel estación/horario y convertimos a formato fecha/hora:

df = df %>%
  group_by(FECHA, DESDE, HASTA, LINEA, ESTACION) %>%  # evaluar agrupar a nivel linea y quedarse con una única línea para el análisis
  summarize(pasajeros = sum(pax_TOTAL)) %>%
  mutate(ts = dmy_hms(paste0(FECHA," " ,DESDE))) %>%
  ungroup() %>%
  select(-c(FECHA, DESDE, HASTA))  ## ver si dejar esto o no

Agrupamos la información a nivel estación/horario y convertimos a formato fecha/hora:

df %>%
  filter(ts > "2022-01-15" & ESTACION == "Florida") %>%
  ggplot(aes(ts, pasajeros)) + geom_line()

Error in `filter()`:
! Problem while computing `..1 = ts > "2022-01-15" & ESTACION == "Florida"`.
Caused by error in `ts > "2022-01-15"`:
! comparison (6) is possible only for atomic and list types
Backtrace:
 1. ... %>% ggplot(aes(ts, pasajeros))
 4. dplyr:::filter.data.frame(., ts > "2022-01-15" & ESTACION == "Florida")
 5. dplyr:::filter_rows(.data, ..., caller_env = caller_env())
 6. dplyr:::filter_eval(dots, mask = mask, error_call = error_call)
 8. mask$eval_all_filter(dots, env_filter)

Tomo estación Palermo para todo el período de análisis

df_analisis <- df_linea %>%
  filter(ts > "2022-08-08" & ts < "2022-08-15" & ESTACION == "Palermo")

df_analisis_hora <- df_analisis %>% 
  mutate('Hora'=hour(ts),'Dia'=date(ts)) %>% 
  mutate('Hora'=paste0(Hora,':00'))

df_analisis_hora <- df_analisis_hora %>% 
  group_by(LINEA,ESTACION,Dia,Hora) %>% 
  summarise(pasajeros = sum(pasajeros))

`summarise()` has grouped output by 'LINEA', 'ESTACION', 'Dia'. You can override using the `.groups` argument.

df_analisis_hora <- df_analisis_hora %>% 
  mutate(ts = parse_date_time(paste0(Dia," ",Hora), "Ymd HMS", truncated = 3))

################################  

df_analisis_hora_test <- df_linea %>%
  filter(ts > "2022-08-15" & ts < "2022-09-01" & ESTACION == "Palermo")

df_analisis_hora_test <- df_analisis_hora_test %>% 
  mutate('Hora'=hour(ts),'Dia'=date(ts)) %>% 
  mutate('Hora'=paste0(Hora,':00'))

df_analisis_hora_test <- df_analisis_hora_test %>% 
  group_by(LINEA,ESTACION,Dia,Hora) %>% 
  summarise(pasajeros = sum(pasajeros))

`summarise()` has grouped output by 'LINEA', 'ESTACION', 'Dia'. You can override using the `.groups` argument.

df_analisis_hora_test <- df_analisis_hora_test %>% 
  mutate(ts = parse_date_time(paste0(Dia," ",Hora), "Ymd HMS", truncated = 3))

  

summary(df_linea)

    LINEA             ESTACION           pasajeros            ts                        
 Length:196190      Length:196190      Min.   :   1.0   Min.   :2022-08-01 05:15:00.00  
 Class :character   Class :character   1st Qu.:  31.0   1st Qu.:2022-08-08 17:45:00.00  
 Mode  :character   Mode  :character   Median :  73.0   Median :2022-08-16 15:15:00.00  
                                       Mean   : 117.2   Mean   :2022-08-16 13:15:56.01  
                                       3rd Qu.: 154.0   3rd Qu.:2022-08-24 10:00:00.00  
                                       Max.   :4146.0   Max.   :2022-08-31 23:30:00.00  

p <- ggplot(df_analisis_hora,aes(ts, pasajeros)) + geom_line() + ggtitle('Pasajeros para Palermo en dos semanas')
ggplotly(p)

Chequeo estacionalidad de serie

Test de Dickey-Fuller

Modelo autoregresivo

library(forecast)

auto_arima <- auto.arima(df_analisis_hora[,5])

summary(auto_arima)

Series: df_analisis_hora[, 5] 
ARIMA(2,0,0) with non-zero mean 

Coefficients:
         ar1      ar2      mean
      1.1508  -0.4620  803.9743
s.e.  0.0771   0.0775   68.5223

sigma^2 = 61146:  log likelihood = -900.03
AIC=1808.07   AICc=1808.39   BIC=1819.54

Training set error measures:
                     ME     RMSE      MAE       MPE     MAPE     MASE        ACF1
Training set -0.4661518 244.4081 194.2803 -101.9805 120.8804 0.903825 -0.03374763

valores_prediccion <- forecast(auto_arima,24)
autoplot(valores_prediccion)